Replication study concerning the effects of homeopathic dilutions of histamine on human basophil degranulation in vitro
Various investigators have observed significant effects of highly diluted histamine on human basophil degranulation in vitro, compared to corresponding water controls. However, active and inactive dilution levels differed in most studies. We aimed to reproduce former studies with flow-cytometry using rigorously controlled experimental conditions to minimise confounding factors. In seven independent experiments, basophils of the same human donor were incubated with diluted histamine (up to 10(-34)M) or water controls and activated with anti-IgE antibodies. Basophil activation was determined by using bi-colour flow-cytometry. Experiments were blinded and performed with a randomised arrangement of the solutions on microtiter-plates. Histamine at the dilutions 10(-2)M and 10(-22)M was associated with a significant inhibition of basophil degranulation (p=0.018, Wilcoxon signed rank test) of 23.1% and 5.7%, respectively, if compared to "diluted" water treated in an identical manner. However, if all controls were pooled, only histamine 10(-2)M had a significant effect. Significant effects were seen for row numbers of the microtiter plates. We were not able to confirm the previously reported large effects of homeopathic histamine dilutions on basophil function of the examined donor. Seemingly, minor variables of the experimental set up can lead to significant differences of the results if not properly controlled.
Complementary Therapies in Medicine (2005) 13, 91—100
Replication study concerning the effects of
homeopathic dilutions of histamine on human
basophil degranulation in vitro
Adrian G. Guggisberg
, Stephan M. Baumgartner
Cornelia M. Tschopp
, Peter Heusser
Institute for Complementary Medicine (KIKOM), University of Berne, Imhoof-Pavillon, 3010 Berne,
Institute for Immunology und Allergology, Inselspital, University of Berne, 3010 Berne, Switzerland
Institute Hiscia, Kirschweg 9, 4144 Arlesheim, Switzerland
Background: Various investigators have observed signiﬁcant effects of highly diluted
histamine on human basophil degranulation in vitro, compared to corresponding
water controls. However, active and inactive dilution levels differed in most studies.
Objective: We aimed to reproduce former studies with ﬂow-cytometry using rigor-
ously controlled experimental conditions to minimise confounding factors.
Methods: In seven independent experiments, basophils of the same human donor
were incubated with diluted histamine (up to 10
M) or water controls and acti-
vated with anti-IgE antibodies. Basophil activation was determined by using bi-
colour ﬂow-cytometry. Experiments were blinded and performed with a randomised
arrangement of the solutions on microtiter-plates.
Results: Histamine at the dilutions 10
M and 10
M was associated with a signif-
icant inhibition of basophil degranulation (p = 0.018, Wilcoxon signed rank test) of
23.1% and 5.7%, respectively, if compared to ‘‘diluted’’ water treated in an identical
manner. However, if all controls were pooled, only histamine 10
M had a signiﬁcant
effect. Signiﬁcant effects were seen for row numbers of the microtiter plates.
Conclusion: We were not able to conﬁrm the previously reported large effects of
homeopathic histamine dilutions on basophil function of the examined donor. Seem-
ingly, minor variables of the experimental set up can lead to signiﬁcant differences
of the results if not properly controlled.
© 2005 Elsevier Ltd. All rights reserved.
Funding: Support was provided from institutional sources only.
* Corresponding author. Tel.: +41 31 632 97 58; fax: +41 31 632 42 62.
E-mail address: email@example.com (A.G. Guggisberg).
0965-2299/$ — see front matter © 2005 Elsevier Ltd. All rights reserved.
92 A.G. Guggisberg et al.
Human basophils are a rare granulocyte type
accounting for 0.1—1% of peripheral blood leuko-
cytes. This cell type has been extensively used to
study the effects of highly diluted substances in
vitro as applied in homeopathy and anthroposophi-
cal medicine. All studies are based on the fact that
human basophils can be activated by anti-human-
This leads to a degranulation
process consisting of a fusion of cytoplasmatic gran-
ules with the plasma membrane and the release of
inﬂammatory mediators, such as histamine.
Different procedures can be used for measuring
basophil activation: (1) By means of staining and
microscopical counting of basophils, degranulated
(i.e. activated) cells can directly be visualised and
be put in relation to brightly stained (i.e. inactive)
However, even if this method is used in
a blinded manner, it may be prone to subjective
inﬂuences. (2) Histamine release of activated
basophils can be determined using ﬂuorimetric
assays of the histamine content in the surrounding
media compared to histamine content of the
(3) Newer ﬂow-cytometric methods allow
the measurement of basophil degranulation using
larger cell numbers.
The fusion of cytoplasmatic
granules of activated basophils with the plasma
membrane leads to the expression of the marker
CD63 on the cellular surface. The percentage of
CD63 expressing basophils can be determined with
ﬂow-cytometry and correlates linearly with the
histamine release of basophils.
Table 1 gives an overview of published studies on
this subject and the methodologies used.
In 1988, a research group around J. Ben-
veniste published their observations of a signiﬁ-
cant basophil degranulation induced by anti-IgE at
dilutions up to a level of 10
as assessed by
This publication provoked
an intensive correspondence. A team led by the
publishing chief editor demonstrated that the vari-
ability of the positive results was smaller than the
statistically expected error of microscopical cell
count and found no effect of higher dilutions, when
the experiments were blinded.
Whereas the ini-
tial ﬁnding were later replicated by Benveniste’s
concluded to be
unable to reproduce these positive results.
Histamine, one of the inﬂammatory mediators
released by basophils, can bind to H2 receptors on
the surface of basophil granulocytes, when present
in high concentrations (>10
M). Thus, it regulates
the basophil degranulation by exerting a feedback
This ﬁnding led others to examine
the effect of very high dilutions of histamine on
basophils stimulated by anti-IgE.
All of the
three methodologies described above were used
to assess the extent of basophil activation. As can
be seen in Table 1, blinding of the experiments
was only systematically performed in studies using
visual determination of basophil activation. In the
initial ﬂow-cytometric studies of the effects of high
, anti-IgE antibodies were used
to gate basophils. This method has the disadvan-
tage that the labelling antibodies have a possibly
confounding activating effect on basophils.
problem has recently been overcome by a new ﬂow-
cytometric technique which identiﬁes basophils by
gating CD123 positive and CD2, CD14, CD16, CD19,
HLA-DR negative cells.
The aim of our study was to perform an indepen-
dent replication of former studies, which used ﬂow-
cytometry for determining basophil activity.
We applied a rigorous protocol to minimize possible
confounding factors. In particular, all experiments
were performed with a blinded and randomised
arrangement of dilutions and controls.
Material and experimental procedures were
adapted from Sainte-Laudy and Belon
in order to be able to compare the
results. HEPES—EDTA buffer (1×), contained
127 mM NaCl (7.42 g/l), 5 mM KCl (0.37 g/l),
20 mM HEPES (Sigma, 4.76 g/l), 5 mM EDTA (Sigma,
1.46 g/l), 5000 IU/l Heparin (Liquemin
adjusted to pH 7.4 with NaOH. HEPES—calcium
Buffer, pH 7.4, contained 127 mM NaCl, 5 mM KCl,
20 mM HEPES, 20 mM CaCl
O (2.94 g/l), 5 mM
O (1.02 g/l). Histamine dihydrochlo-
ride and formyl-methionyl-leucyl-phenylalanine
(fMLP) was purchased from Sigma and dextran
was obtained from Pharmacia Biotech. Rabbit-
anti-human IgE used to activate basophils was
from DAKO. FITC-anti-IgE and PE-anti-CD63 used
for ﬂow-cytometry were from Caltag. FITC-
anti-IgG and PE-IgG1 (Caltag) served as staining
Preparation of histamine and sham dilutions
A 0.1-M solution was prepared by adding 184 mg
histamine to 10 ml distilled water. The 10
tion was obtained by 10-fold dilution of histamine
0.1 M with distilled water, followed by 10 s of vor-
texing at maximal speed. Concentrations 10
Replication study concerning the effects of homeopathic dilutions 93
Table 1 Overview of published studies investigating the effects of homeopathic solutions on human basophils and the methodologies used.
Study Method Pre-treatment Homeopathic
Measured parameter Blinding
Davenas et al.
Visual counting — Anti-IgE % Degranulated basophils Some of the
Maddox et al.
Visual counting — Anti-IgE % Degranulated basophils Yes
Poitevin et al.
Visual counting — Lung histamine,
% Degranulated basophils Yes
Benveniste et al.
Visual counting — Anti-IgE, Apis
% Degranulated basophils Yes
Ovelgonne et al.
Visual counting — Anti-IgE % Degranulated basophils Yes
Hirst et al.
Visual counting — Anti-IgE % Degranulated basophils Yes
Visual counting — Histamine % Inhibition of basophil
— Histamine % Inhibition of basophil
Histamine Basophil CD63 expression No
Belon et al.
Visual counting — Histamine % Inhibition of basophil
% Activation of basophil
— Histamine % Inhibition of basophil
Lorenz et al.
CD2−, CD14−, CD16−,
Histamine % Inhibition of basophil
Lorenz et al.
CD2−, CD14−, CD16−,
Ficoll isolation Histamine Test stability, basophil
Belon et al.
Visual counting — Histamine % Degranulated basophils Yes
Histamine Basophil CD63
expression, % inhibition
of basophil CD63
of histamine in supernatant
and in cells
— Histamine Histamine release No
94 A.G. Guggisberg et al.
M were prepared by serial 100-fold dilution
with distilled water, followed by vortexing for 10 s
at maximum speed. New pipette tips and tubes
were used for each dilution step. Sham solutions
(‘‘diluted water’’) were analogously prepared by
serial dilution and vortexing of distilled water. His-
tamine and sham concentrations 10
M were made isotonic by adding 1 part in
10 HEPES—EDTA buffer in a 10-fold concentration.
Solutions were stored at 4
C for 1—6 days.
Preparation of leucocytes
Thirty millilitres of fresh blood anticoagulated with
EDTA from healthy non-allergic volunteers (after
informed consent according to the declaration of
Helsinki) was mixed (in the main experiments)
with 7.5 ml dextran 6%. After sedimentation of
erythrocytes for 90 min at room temperature, the
leukocyte-rich plasma was collected and washed
with HEPES—EDTA buffer and adjusted to a concen-
tration of 70—100 million cells per millilitres.
Twenty millilitres of leukocyte suspension was
aliquoted in the wells of a 96-microtiter plate,
mixed with 20 l of buffer control, histamine or
sham solutions and incubated during 30 min at room
temperature. Then, 20 l anti-IgE (ﬁnal concentra-
tion 0.2 g/ml) in HEPES—calcium was added to the
cells followed by incubation at 37
C for 30 min.
EDTA-buffer (‘‘untreated water’’), calcium buffer
and fMLP (ﬁnal concentration 8.3 × 10
as negative and positive controls. The reaction was
stopped by adding cooled HEPES—EDTA buffer. After
centrifugation (200 × g and 4
C, 10 min), the cells
were labelled in a volume of 50 l with 1 g FITC-
anti-IgE and 1 g PE-anti-CD63 or control antibod-
ies. After 20 min at 4
C, the cells were washed
with HEPES—EDTA buffer and analysed by ﬂow-
Flow-cytometry was performed using Becton Dick-
inson FACSCalibur 4FL. Lymphocytes and basophils
were gated according to their distribution in the
FSC/SSC dot plot. From this, basophils were gated
by their bright anti-IgE FITC ﬂuorescence. Two
hundred to ﬁve hundred basophils were counted,
depending on the cell concentration available. The
percentage of active basophils was calculated by
setting an electronic gate between CD63+ and
CD63− basophils corresponding to 1—2% positive
cells stained with the PE-isotype matched control
During pilot experiments, the inﬂuence of the
methodological set-up on basophil activation was
assessed. No blinding was performed for pilot
experiments. A possible difference of basophil
responses when using either tubes, culture plates
or microtiter-plates was investigated by comparing
the results obtained with otherwise equally treated
basophils of one donor. The inﬂuence of dex-
tran sedimentation was studied in three subjects.
The optimal amount of activating anti-IgE was
tested by establishing a dose—effect curve for ﬁve
For the principal experiments, precautions were
taken in order to minimise effects due to factors
other than homeopathic dilutions of histamine.
Seven independent experiments were performed
with the blood of one single donor to avoid inter-
individual differences. This donor was chosen
on account of the sufﬁciently high proportion
of basophils in his leucocyte suspensions (mean
1.16 ± 0.6%) and because his basophils showed the
desired moderate response to activation with anti-
IgE antibodies (mean activation 30.16 ± 10.94%).
The response to low histamine concentrations was
not previously tested and was not a criteria for
the selection of this donor. Histamine dilutions
were compared with sham solutions (see above).
Histamine and sham solutions were blinded with a
letter code. All coded solutions were performed in
triplicates instead of duplicates used in most stud-
ies. Experiments were done on 96-well plates with a
computer-randomised arrangement for the position
of each experimental condition, including the repli-
cates, and this random arrangement was newly gen-
erated for each experiment. The outer wells of the
plate were not used in order to avoid possible bor-
der effects. Minimisation of differences in delays of
incubation was achieved by transferring the coded
solutions and activators from mirror plates to the
experimental plate using multi-pipettes, and by
limiting the number of concentrations tested. The
M and 10
selected for our study, since they had shown the
most pronounced effects in previous investigations
(see Tables 1 and 3).
Replication study concerning the effects of homeopathic dilutions 95
Table 2 Mean activation as determined by % CD63 positive basophils is shown for all seven independent experi-
ments and every experimental condition (+: histamine and H
O were made isotonic by adding EDTA buffer).
Pre-incubation Activation Experiment number Mean S.D.
Buffer Buffer 3.91 0.96 0.55 1.45 1.62 2.96 4.57 2.29 1.54
Buffer aIgE 45.65 30.95 20.49 19.66 24.55 24.95 44.84 30.16 10.94
Buffer fMLP 50.843.89 36.94 40.44 40 57.82 68.29 48.31 11.37
Histamine+ aIgE 42.05 21.62 13.21 14.84 22.84 28.08 31.84 24.93 10.04
O+ aIgE 48.727.09 28.97 17.76 24.77 37.35 41.96 32.37 10.76
Difference −13.66 −20.19 −54.40 −16.44 −7.79 −24.82 −24.12 −23.06 15.06
Histamine+ aIgE 47.04 28.82 28.92 17.32 23.69 27.32 38.73 30.26 9.80
O+ aIgE 50.99 30.928.27 15.48 26.53 27.336.95 30.92 10.93
Difference −7.75 −6.73 2.30 11.89 −10.70 0.074.82 −0.87 8.00
Histamine+ aIgE 49.430.67 24.33 20.87 22.55 32.87 43.76 32.06 10.92
O+ aIgE 45.85 27.52 24.27 22 24.84 34.74 39.83 31.29 9.02
Difference 7.74 11.45 0.25 −5.14 −9.22 −5.38 9.87 1.37 8.32
Histamine+ aIgE 46.931.221.03 18.32 24.59 32.08 37.84 30.28 9.99
O+ aIgE 50.63 31.27 22.519.328.16 33.46 39.25 32.08 10.56
Difference −7.37 −0.22 −6.53 −5.08 −12.68 −4.12 −3.59 −5.66 3.86
Histamine+ aIgE 53.54 28.92 29.18 20.24 23.93 39.55 42.66 34.00 11.74
O+ aIgE 43.435.29 27.16 19.69 25.829.33 40.731.62 8.52
Difference 23.36 −18.05 7.44 2.79 −7.25 34.84 4.82 6.85 17.78
Histamine+ aIgE 49.26 29.77 23.72 22.15 22.08 34.29 37.58 31.26 10.00
O+ aIgE 53.36 33.77 22.79 16.61 24.26 34.33 38.12 31.89 12.14
Difference −7.68 −11.84 4.0833.35 −8.99 −0.12 −1.42 1.0615.29
Histamine+ aIgE 46.94 31.82 27.86 17.97 24.91 38.46 41.05 32.72 10.06
O+ aIgE 47.53 31.69 24.28 18.37 26.67 31.63 42.17 31.76 10.15
Difference −1.24 0.41 14.74 −2.18 −6.60 21.59 −2.66 3.44 10.47
Histamine+ aIgE 51.75 30.94 26.38 17.62 21.15 32.38 43.94 32.02 12.18
O+ aIgE 48.83 29.18 27.516.41 17.53 36.34 40.36 30.88 11.86
Difference 5.98 6.03 −4.07 7.37 20.65 −10.90 8.87 4.85 10.02
The difference between histamine and water dilution was calculated by the formula (W
, where W was the percent
activitation of basophils pre-incubated with distilled water; H, percent activation of basophils pre-incubated with histamine and
x the corresponding dilution step.
96 A.G. Guggisberg et al.
Table 3 Comparison of the results of the ﬁve previous studies using ﬂow-cytometry to assess the effect of home-
opathic histamine on human basophils and our investigation.
Belon et al.
6 n.s. 43
n.s. 22 18 6
13 n.s. 14 18
13 n.s. −11
2 n.s. 1 −8 −3
26 14 −1
17 10 −5
2 n.s. −12 3
14 n.s. 16 9
12 n.s. 8 13
Two further studies are not tabulated since they contain preliminary results only.
Mean inhibition of basophil CD63 expression
by centesimal dilutions of histamine is indicated in percent compared to the corresponding water control (n.s.: not signiﬁcant,
no mean values reported). Data were estimated from graphs if not explicitly quoted.
p < 0.05
p < 0.01.
p < 0.001.
Analysis of the data
The main hypothesis (‘‘there is no difference
between the CD63-expression of basophils incu-
bated with histamine and sham solutions, i.e.
diluted water’’) was statistically assessed for each
pair of dilutions using the non-parametric Wilcoxon
signed rank test (independent planned compar-
isons, therefore, without Bonferroni correction
In an additional analysis a posteriori, degranu-
lation of basophils incubated with histamine was
compared to the data of the EDTA-buffer control
(‘‘untreated water’’) of the corresponding exper-
iment and to the mean value of all sham solu-
tions of the corresponding experiment. In this case,
a Bonferroni correction was applied to the non-
parametric Wilcoxon signed rank test since all
homeopathic dilutions were compared simultane-
ously to one control.
Analysis of variance (F-test) was used for the sta-
tistical test of plate-effects: data were analysed for
border-, row-, and column-effects. Columns were
labelled with the letters B to G on 96-well plates
and handled simultaneously with multi-pipettes,
whereas rows were labelled with the numbers 2—11
and handled subsequently. Environmental effects
(inﬂuence of temperature and age of the dilutions
on basophil degranulation) were assessed by an
exploratory data analysis (F- and t-tests).
Inﬂuence of methodological parameters on
The usage of polystyrene tubes, cell culture plates
or conventional microtiter plates did not signif-
icantly inﬂuence basophil responses (results not
No signiﬁcant difference was found between
degranulation of anti-IgE activated basophils
obtained with or without dextran sedimentation. In
an exploratory investigation, no major inﬂuence of
dextran on the effect of histamine was observed.
However, signiﬁcantly more basophils could be
obtained when dextran was used (mean increase
of 289%). Therefore, in the principal experiments,
erythrocyte-sedimentation was accelerated with
Replication study concerning the effects of homeopathic dilutions 97
dextran. Dose—response curves of anti-IgE revealed
that a concentration 0.2 g/ml, which was also
used in previous work of other groups,
a near maximal effect. Therefore, this concentra-
tion was used for our experiments.
Effect of histamine dilutions on CD63 expression
in basophils activated with anti-IgE
As shown in Table 2, CD63-expression was increased
by anti-IgE (mean 30.16 ± 10.94%) and fMLP (mean
48.31 ± 11.37%) as compared to the buffer control
(mean 2.29 ± 1.54%).
The effect of pre-incubating the basophils with
different dilutions of histamine or water treated in
an identical manner is shown in Fig. 1 and was ana-
lyzed in three different manners. The black bars
show the effect of pre-incubation with histamine
at the indicated concentration when compared
to the corresponding concentration of ‘‘diluted
water’’. A pharmacological concentration of his-
M) inhibited basophil degranulation
by 23% (p = 0.018, Wilcoxon signed rank test), in
agreement with former studies. A rather weak,
but statistically signiﬁcant, inhibition of 5.7% was
found for the homeopathic concentration of 10
(p = 0.018, Wilcoxon signed rank test). None of the
other homeopathically diluted histamine concen-
trations showed a signiﬁcant effect.
A posteriori, the same data set was also anal-
ysed by comparing the effect of histamine to the
corresponding ‘‘untreated water’’ control. In this
Figure 1 Inhibition of basophil degranulation (expressed
as percentage of the corresponding control) observed
after incubation of basophils with dilutions of his-
tamine (mean ± standard error ± double standard error,
p < 0.05). Black bars: histamine vs. corresponding
diluted water control; shaded bars: histamine vs.
untreated water control; open bars: histamine vs. the
pool of all treated water controls.
case, none of the homeopathic histamine dilutions
showed a statistically signiﬁcant effect (Fig. 1,
shaded bars). In a third analysis, all ‘‘diluted
water’’ controls of one experiment were averaged
and compared to the effect of the different his-
tamine dilutions. Again, in contrast to the pharma-
cological histamine concentration, highly diluted
solutions had no signiﬁcant effect (Fig. 1, open
Microtiter plate position effects
Microtiter-plate effects are a quite common phe-
nomenon. We, therefore, analysed the raw data of
the main experiments for column, row and border
effects. Data of controls and pharmacological con-
centrations of histamine were excluded from this
Overall analysis of variance reveals no signiﬁ-
cant column (p > 0.05) or border (p > 0.4) effect,
but a signiﬁcant row effect (p < 0.0001). Overall,
CD63-expression decreased for higher row num-
bers with an increment of −0.6% per row number.
The row number effect additionally shows a signiﬁ-
cant interaction with the number of the experiment
(p < 0.0001). Linear regression between basophil
activation and row number yields a statistically sig-
niﬁcant correlation for four single experiments: a
decrease for three experiments (p < 0.05, p < 0.001
and p < 0.001, respectively) and an increase for one
experiment (p < 0.002).
As expected, randomisation effectively elimi-
nated these microtiter-plate effects. A compari-
son of histamine and water dilutions did not yield
any signiﬁcant difference (p > 0.05) for the row
number, neither for all dilution levels pooled nor
A closer inspection of Fig. 1 reveals an increased
standard deviation for dilution levels <10
therefore, performed an exploratory data analysis
in order to determine possible modulating factors.
The experiments were performed in summer
time with considerable variation of the room
temperature, which was registered for each
experiment. Whereas the temperature during the
preparation of the dilutions does not seem to be
an important factor, a high room temperature
during incubation of basophils with histamine at a
dilution level of 10
M compared to controls was
associated with a signiﬁcant (p < 0.05) increase of
CD63 expression (negative inhibition). However, for
all other dilution levels, no signiﬁcant differences
can be found between any pair of temperature
(p > 0.05) and a two-way analysis of variance of
the pure differences between histamine and water
98 A.G. Guggisberg et al.
dilutions does not yield a signiﬁcant interaction
between dilution level and incubation temperature
(p > 0.3).
Histamine- and sham-dilutions were stored at
C for 1—6 days before usage. A two-way analysis
of variance does not yield a signiﬁcant interaction
between dilution level and storage time on the pure
differences between histamine and water dilutions
(p > 0.1). In particular, no signiﬁcant differences
can be found between a storage time of 1 day com-
pared to storage times of 2—6 days (p > 0.1) for
the pharmacological concentration of 10
cating that molecular histamine was stable when
stored at 4
C. However, a qualitative comparison
of a storage of 1 day to storage times of 2—6 days
could suggest, that homeopathic histamine dilu-
tions below 10
M enhance anti-IgE induced CD63-
expression when prepared only 1 day before use, in
contrast to the previously reported inhibitory effect
of homeopathic histamine. A t-test for the dilution
M yields a signiﬁcant difference for a
storage time longer than 1 day (p < 0.002).
Several studies have reported signiﬁcant effects
of highly diluted solutions on basophil degranula-
tion in vitro, using different rationales and differ-
ent methodologies for the assessment of basophil
activation (Table 1). The initially reported signiﬁ-
cant activation of basophil degranulation by home-
opathic dilutions of activating anti-IgE-antibodies
could not be reproduced when experiments were
blinded and properly controlled.
very high dilutions of histamine have shown signiﬁ-
cant effects on basophil activation in all published
studies so far.
When assessing basophil degranulation by
visual counting of stained basophils, a signif-
icant inhibition of basophil degranulation by
histamine was found in some of the tested very
Furthermore, an overall
inhibitory effect of highly diluted histamine was
conﬁrmed by a European multicentre study.
However, the results differed to a large amount
between the four laboratories involved, which
was ascribed to inter-individual differences of the
As can be seen in Table 3, several studies
using ﬂow-cytometry to measure basophil activa-
tion observed a signiﬁcant inhibitory or activating
effect of histamine at some of the tested home-
opathic concentrations. However, active and inac-
tive dilution levels differed in most investigations.
For example, a concentration of 10
M caused a
signiﬁcant inhibition in one study,
whereas a sig-
niﬁcant activation with the same concentration was
(see Table 3).
In a more recent publication of Belon et al.,
signiﬁcant effects of highly diluted histamine were
also found when basophil activation was measured
with the histamine release method. However, these
experiments as well as most of the ﬂow-cytometric
studies do not state to have performed blinding of
diluted histamine and controls. The only exception
are the recent studies of Lorenz et al.,
contain only preliminary results.
In this study, we tried to reproduce the repeat-
edly reported inhibitory effect of highly diluted
histamine solutions on anti-IgE induced basophil
degranulation using the ﬂow-cytometric method
developed by Sainte-Laudy et al.
The blinding of
the solutions and the (computer generated) ran-
domisation of the experimental set-up including the
position of replicates on the microtiter-plate by
far exceeded the usual precautions to avoid arte-
facts in ‘‘conventional’’ in vitro research. In con-
trast to former studies (Table 3), no large effect
of highly diluted histamine solutions on anti-IgE
induced basophil degranulation as assessed by CD63
up-regulation can be found in our data under these
conditions. But notably, even with this rigid pro-
tocol we found one minor, but statistically signif-
icant, inhibitory effect at a histamine dilution of
M, when compared to the effect of water
‘‘diluted’’ to 10
M. However, when the same
data were compared a posteriori to other reason-
able controls, statistical signiﬁcance was lost. One
may argue, that statistical signiﬁcance was reached
only because the standard deviation of the inhibi-
tion within this dilution level (10
M) happened to
be small (Table 2 and Fig. 1). Since the p-value is
rather modest (p = 0.018), the evidence for reject-
ing the null hypothesis (‘‘no difference between
histamine and water dilutions’’) is not overwhelm-
ing, i.e. it is possible that this result is due to
One interest of our study was to raise hypothe-
ses, which might explain the differing results in
previous work and in our investigation (Table 3).
Whereas at pharmacological concentrations the
results were quite similar in the different stud-
ies, larger differences were found at higher dilu-
tions, especially at 10
M, where inhibitory as
well as activating effects were described. Four dif-
ferent explanations can be assumed: (1) A differing
methodology might provoke divergent results. (2)
The study results might depend on inter-individual
differences of blood donors. (3) Unidentiﬁed con-
founding parameters might provoke a systematic
error and even false positive results, which mimic
Replication study concerning the effects of homeopathic dilutions 99
an effect of histamine at high dilutions. (4) External
environmental factors could inﬂuence the experi-
(1) Methodological differences among previous
studies might be responsible for the diver-
gent results. Indeed, it has been suggested
that the dilution medium may inﬂuence the
effects of high dilutions on basophils.
ever, identical solutions and antibodies were
mostly used and the methodology developed by
Sainte-Laudy and Belon
was followed closely
in previous reports and in our study. In a pilot
experiment, we ruled out a possible inﬂuence
of using microtiter-plates (as used in
) or sin-
gle polystyrene tubes (used in
). Unlike other
we found an acceleration of
leukocyte sedimentation with dextran to be
necessary. In pilot experiments, we ascertained
that this difference had no signiﬁcant inﬂu-
ence on basophil degranulation and on effect of
substantial histamine. A negative inﬂuence of
dextran on effects of highly diluted histamine
formally cannot be excluded, though.
It is noteworthy, that dilutions in our and
most previous studies using vortex instead of
long lasting shaking were not produced accord-
ing to homeopathic recommendations. How-
ever, major effects were reported with this
technique (Table 3).
(2) One may argue that the optimally active home-
opathic dilution may differ for cells isolated
from different individuals, obscuring an effect
when the different experiments are combined.
In order to avoid this objection, the blood of
one single donor was used for all our seven
main experiments. On the other hand, it is also
possible that the basophils of this single donor
investigated are not or only weakly susceptible
to homeopathic dilutions of histamine, explain-
ing the smaller effects observed in our study.
(3) We observed highly signiﬁcant effects related
to the microtiter-plate position, although maxi-
mal efforts were taken to avoid such inﬂuences.
We do not have a convincing explanation for this
effect. Since row effects were observed, differ-
ences of the incubation time would seem most
probable. However, by using multi-pipettes and
mirror-plates, and by controlling the incuba-
tion times of each row with a clock, the dif-
ferences in incubation time were maximally
60 s, which seems to be too small to cause
larger effects. These difﬁculties illustrate the
importance of randomisation and blinding of
a sufﬁcient number of replicates within each
experiment, which effectively eliminated con-
founding plate effects in our study. Lacking
randomisation might yield false positive results
and might be responsible for the varying results
among different investigations.
In this study, experimental procedures were
adapted from earlier studies,
large effects of high histamine dilutions were
reported. However, by using improved buffer
isolation of leucocytes
basophils as well as by assessing basophil
activation with an improved ﬂow-cytometric
or measurement of histamine
a more stable experimental system
might be obtained in future studies.
(4) A closer inspection of Fig. 1 reveals an
increased standard error for dilution levels
M. An exploratory data analysis gives
rise to the hypothesis, that the age of the his-
tamine dilutions and the temperature during
the incubation of the basophils might modulate
their response to homeopathic dilutions of his-
Additional studies using strictly controlled
experimental conditions, several blood donors and
an improved methodology are needed to elucidate
the open questions.
We would like to thank Professor M. Ennis and V.
Brown, Queens University of Belfast, Professor C.A.
Dahinden, University of Berne, and D. Shah, Insti-
tute Hiscia, for their kind help with methodological
1. Nakagawa T, Stadler BM, de Weck AL. Flow-cytometric
analysis of human basophil degranulation. I. Quantiﬁca-
tion of human basophils and their degranulation by ﬂow-
cytometry. Allergy 1981;36:39—47.
2. Gilbert HS, Ornstein L. Basophil counting with a new stain-
ing method using alcian blue. Blood 1975;46:279—86.
3. Shore PA, Burkhalter A, Cohn Jr VH. A method for the ﬂuori-
metric assay of histamine in tissues. J Pharmacol Exp Ther
4. Knol EF, Mul FPJ, Jansen H, Calafat J, Roos D. Monitoring
human basophil activation via CD63 monoclonal antibody
435. J Allergy Clin Immunol 1991;88:328—38.
5. Sainte-Laudy J, Sabbah A, Vallon C, Gu
erin JC. Analysis
of anti-IgE and allergen induced human basophil activa-
tion by ﬂow cytometry. Comparison with histamine release.
Inﬂamm Res 1998;47:401—8.
6. Davenas E, Beauvais F, Amara J, Oberbaum M, Robin-
zon B, Miadonna A, et al. Human basophil degranulation
triggered by very dilute antiserum against IgE. Nature
100 A.G. Guggisberg et al.
7. Maddox J, Randi J, Stewart WW. ‘‘High-dilution’’ experi-
ments a delusion. Nature 1988;334:287—91.
8. Benveniste J, Davenas E, Ducot B, Cornillet B, Poitevin B,
Spira A. L’agitation de solutions hautement dilu
e biologique sp
eciﬁque. CR Acad Sci Paris
9. Ovelgonne JH, Bol AW, Hop WC, van Wijk R. Mechanical agi-
tation of very dilute antiserum against IgE has no effect on
basophil staining properties. Experientia 1992;48:504—50.
10. Hirst SJ, Hayes NA, Burridge J, Pearce FL, Fore-
man JC. Human basophil degranulation is not triggered
by very dilute antiserum against human IgE. Nature
11. Lichtenstein LM, Gillespie E. Inhibition of histamine
release by histamine controlled by H2 receptor. Nature
12. Tung R, Kagey-Sobotka A, Plaut M, Lichtenstein LM.
H2 antihistamines augment antigen-induced histamine
release from human basophils in vitro. J Immunol
13. Poitevin B, Davenas E, Benveniste J. In vitro immunolog-
ical degranulation of human basophils is modulated by
lung histamine and Apis melliﬁca. Br J Clin Pharmacol
14. Sainte-Laudy J, Belon P. Inhibition of human basophil
activation by high dilutions of histamine. Agents Actions
15. Sainte-Laudy J, Belon P. Analysis of immunosuppressive
activity of serial dilutions of histamine on human basophil
activation by ﬂow cytometry. Inﬂamm Res 1996;45(Suppl.
16. Sainte-Laudy J, Belon P. Application of ﬂow cytometry to
the analysis of the immunosuppressive effect of histamine
dilutions on human basophil activation: effect of cimeti-
dine. Inﬂamm Res 1997;46(Suppl. 1):S27—8.
17. Belon P, Cumps J, Ennis M, Mannaioni PF, Sainte-Laudy J,
Roberfroid M, et al. Inhibition of human basophil degranula-
tion by successive histamine dilutions: results of a European
multi-centre trial. Inﬂamm Res 1999;48(Suppl. 1):S17—8.
18. Sainte-Laudy J. Stimulatory effect of high dilutions of his-
tamine on activation of human basophils induced by anti-
IgE. Inﬂamm Res 2001;50(Suppl. 2):S63—4.
19. Brown V, Ennis M. Flow-cytometric analysis of basophil
activation: inhibition by histamine at conventional and
homeopathic concentrations. Inﬂamm Res 2001;50(Suppl.
20. Lorenz I, Schneider EM, Stolz P, Brack A, Strube J. Inﬂuence
of the diluent on the effect of highly diluted histamine on
basophil activation. Homeopathy 2003;92:11—8.
21. Lorenz I, Schneider EM, Stolz P, Brack A, Strube J. Sensitive
ﬂow cytometric method to test basophil activation inﬂu-
enced by homeopathic histamine dilutions. Forsch Komple-
armed Klass Naturheilkd 2003;10:316—24.
22. Belon P, Cumps J, Ennis M, Mannaioni PF, Roberfroid
M, Sainte-Laudy J, et al. Histamine dilutions modulate
basophil activation. Inﬂamm Res 2004;53:181—8.
23. Hsu JC. Multiple comparisions: theory and methods. New
York: Chapman & Hall/CRC; 1996.